In the present year it was clear that rain had fallen since the locust invasion, and although much damage had been done, tolerable specimens of many plants here seen for the first time were to be found. A few of these are common to the Canary Islands and this part of Africa; others are not yet known except on this coast. The most curious of them is the _Senecio_ (_Kleinia_) _pteroneura_, whose succulent almost leafless branches, as thick as a man’s finger, bear a few heads of flowers that differ little, save in their larger size, from those of the common groundsel. Well pleased with our first glance at the South Marocco flora, we returned to our comfortable quarters, and spent a pleasant evening in discussing our future movements, and in drawing upon our host’s ample stores of information respecting the country and its inhabitants. We were now for the first time brought into contact with the primitive stock of this part of Africa, one main branch of the Bereber race, which is distinguished by speaking some dialect of the Shelluh (Shleuh) language.[2] The affinity of this people with the Berebers of the Lesser Atlas—including under that name the Kabyles of Algeria, with the Riff tribes of North-west Marocco—has been denied, but does not appear to be open to reasonable doubt. The type is physically the same, excepting among some of the tribes south of the Great Atlas, where the intermixture of Negro blood has introduced new and very diverse elements. The languages now spoken among these tribes doubtless exhibit marked differences, especially to the ear of a foreigner. Jackson long ago denied the relationship between the Shelluh and the Bereber, while Washington, in the paper already quoted, came to a contrary conclusion. It may now be considered as beyond question that the differences between the Shelluh and the Kabyle are merely dialectic.[3] The value of linguistic evidence in ethnological inquiries has of late been questioned by eminent critics, and it must be conceded that such evidence, when it merely rests on lexicographical coincidences, is of less value than when it is derived from grammatical structure; yet, after all deductions, the facts remain to be accounted for, and, in the absence of proof to the contrary, it goes far towards proving community of origin. It must be remembered, that unlettered races are subject to far greater and more rapid changes of dialect than those who preserve in sacred books or popular poetry fixed standards of correct speech; add to this, the chances of error when a traveller, communicating with a native through an interpreter, and contending with sounds unusual to his ear, attempts to form a vocabulary. These causes, acting together, tend to increase the difficulty of recognizing linguistic affinities that really exist.